Classifier to predict security threats in the login access of a Firewall server.
This Random Forest classifier model was developed to detect security threats in the Firewall server of a cyberservices firm based in Cambridge, Massachusetts. The data used came directly from logs of the server's login access, updated daily.
A two-way classifier, the initial model received an accuracy of 0.906 with the test datasets, while the final classifier gives predictions with an accuracy of 1.0.
(This was my first assignment & the first modeling problem I solved; the 0.906 tasted much sweeter as I was only starting out, than the later 1.0 by when I'd figured out it was a simple problem).